Field of the Invention
The present invention relates to an image processing apparatus, a radiation imaging apparatus, an image processing method, and a storage medium.
Description of the Related Art
In recent years, a radiation imaging apparatus that two-dimensionally arranges image sensors (detection elements) made of amorphous silicon or single-crystal silicon and captures a radiation image has been put into practical use. In such a radiation imaging apparatus, as processing for reducing the influence of fixed pattern noise caused by the dark current and the like of the image sensors, dark correction processing of subtracting a dark image obtained in a state without radiation irradiation from an object image is performed.
The fixed pattern noise output is known to vary in accordance with an increase/decrease in the dark current caused by a variation in the operation temperature. For this reason, if the fixed pattern noise varies during the time from dark image capturing to object image capturing, uncorrected fixed pattern noise may remain in the object signal and lead to a loss of the contrast of the object signal, or the object signal may be lost by overcorrection.
For example, in a case in which fixed pattern noise decreases, and overcorrection occurs, a pixel having a negative pixel value smaller than 0 is generated in an image after dark correction. When performing logarithmic conversion in image processing after dark correction, an irreversible information loss may occur in object information upon replacing a region including a negative pixel value with positive values. To prevent an information loss in correction, Japanese Patent Laid-Open No. 2010-12082 proposes a technique of shifting entire image data to the positive side in a case in which the minimum pixel value is negative so that the minimum pixel value changes to 0.
However, in the arrangement for performing analysis on an image basis, as in Japanese Patent Laid-Open No. 2010-12082, even if the same object is captured, the shift amount (correction amount) varies in each image due to the influence of random noise and the like in the apparatus. In a case in which the same object is continuously captured as in fluoroscopy, if the shift amount (correction amount) varies in each frame, and the tone of an output image varies like a flicker, it may exert an effect on image processing after dark correction, resulting in degradation in image quality.
The present invention has been made in consideration of the above-described problem, and provides an image processing technique capable of reducing a variation in the correction amount between in each image and suppressing degradation in quality of an output image while preventing a loss of image information.